ScholarWorks@UMassAmherst (University of Massachusetts Amherst) · 1997 · 33 citations · 11 references
Open access
EngineeringComputer ScienceRobotics EngineeringRobot ControlRobot LearningRoboticsDynamic ControlLegged Robot ControlLocomotion LearningPolicy LearningReinforcement LearningRobot Control ArchitectureAbstract Model ConstructionDiscrete-event SystemsHybrid ControlLearning-based ControlReinforcement Learning For Gait GenerationReinforcement Learning For RoboticsSafety ConstraintsUncertain Systems Modeling
Autonomous robot systems operating in an uncertain environment have to be reactive and adaptive in order to cope with changing environment conditions and task requirements. To achieve this, the hybrid control architecture presented in this paper uses reinforcement learning on top of a Discrete Event Dynamic System (DEDS) framework to learn to supervise a set of basis controllers in order to achieve a given task. The use of an abstract system model in the automatically derived supervisor reduces the complexity of the learning problem. In addition, safety constraints may be imposed a priori, such that the system learns on-line in a single trial without the need for an outside teacher. To demonstrate the applicability of the approach, the architecture is used to learn a turning gait on a four legged robot platform. 1 Introduction Autonomous robot systems operating in an uncertain environment have to be able to cope with new situations and task requirements. Important pr...
11
The control of discrete event systems
Peter J. Ramadge, W.M. Wonham · Proceedings of the IEEE · 1989 · 2.9K citations
Learning to coordinate behaviors
Pattie Maes, Rodney A. Brooks · 1990 · 364 citations